No product in this catalogue is a cell-penetrating peptide. The class is covered here because buyers ask about it, because it is one of the most cited ideas in peptide research, and because its best-known members share a property with catalogue peptides such as LL-37: a high positive charge, with the handling consequences that brings. KPV, a tripeptide with a single lysine, sits at the other end of that scale and is a useful contrast.
A cell-penetrating peptide, usually abbreviated CPP, is a short sequence, typically 5 to 30 residues, that in laboratory cell models crosses the plasma membrane and can carry an attached molecule with it. The class was defined by two discoveries in the late 1980s and 1990s, and it has grown into several hundred reported sequences since. This guide covers the founding members, the classes they fall into, what the uptake literature describes and in which models, and what a laboratory ordering one from a custom synthesis house should expect.
Where the class came from: Tat, 1988 and 1997
In December 1988 two papers in the same issue of Cell reported that the Tat protein of human immunodeficiency virus, added to the medium of cultured cells, was taken up by them. Frankel and Pabo showed uptake of the purified protein and localisation to the nucleus; Green and Loewenstein showed the same for a chemically synthesised version and began mapping which parts of the protein were responsible. The observation was unexpected because proteins do not ordinarily cross membranes.
In 1997 Vivès, Brodin and Lebleu narrowed the activity to a short basic stretch of the protein. A peptide covering residues 48 to 60, GRKKRRQRRRPPQ, translocated into cultured cells within minutes and accumulated in the nucleus. The sequence most often used in later work is residues 47 to 57, YGRKKRRQRRR, which carries six arginines and two lysines in eleven residues. This is the TAT peptide, and it is the reference member of the class.
Penetratin: the third helix of Antennapedia
The second founding member came from a different organism. Derossi, Joliot, Chassaing and Prochiantz reported in 1994 that the third helix of the Antennapedia homeodomain, a 16-residue sequence from a Drosophila transcription factor, crossed the membranes of cultured neurons. The sequence, RQIKIWFQNRRMKWKK, was named penetratin. It differs from TAT in having hydrophobic residues (isoleucine, tryptophan, phenylalanine, methionine) interspersed among its seven basic ones, which places it in a different sub-class.
Derossi and colleagues also reported that the uptake did not depend on the peptide's helical structure and was not blocked at 4 °C, observations that fed a long debate about mechanism which the section on uptake returns to.
Polyarginine and the guanidinium group
Once the arginine-rich character of TAT was recognised, the obvious experiment was to test arginine on its own. Futaki and colleagues reported in 2001 that homopolymers of arginine, and arginine-rich segments from a range of unrelated proteins, were taken up by cultured cells, with chains of around eight residues being the most effective. Lysine homopolymers of the same length were markedly less active. The difference is attributed to the guanidinium group of arginine, which can form two hydrogen bonds with the phosphate and sulphate groups on a cell surface where lysine's amine forms one.
Octa- and nona-arginine, written R8 and R9, are now used as minimal CPPs in their own right. They are also the simplest demonstration that the class is defined by a physical property, charge density, rather than by any evolutionary origin.
The three classes
The usual division, set out in reviews such as Guidotti, Brambilla and Rossi in 2017, is by composition. Cationic CPPs are dominated by arginine and lysine with few hydrophobic residues; TAT and polyarginine are the type examples. Amphipathic CPPs have both a basic face and a hydrophobic face, either along the sequence or when folded as a helix; penetratin is the type example, and designed chimeras such as transportan belong here. Hydrophobic CPPs, the smallest group, have few charges and rely on non-polar residues.
| Peptide | Origin | Sequence | Residues | Approximate net charge | Class | First report |
|---|---|---|---|---|---|---|
| TAT 47-57 | HIV-1 Tat protein, basic domain | YGRKKRRQRRR | 11 | +8 | Cationic | Frankel and Pabo 1988 (protein); Vivès and colleagues 1997 (peptide) |
| Penetratin | Antennapedia homeodomain, helix 3 | RQIKIWFQNRRMKWKK | 16 | +7 | Amphipathic | Derossi and colleagues 1994 |
| Octa-arginine (R8) | Designed homopolymer | RRRRRRRR | 8 | +8 | Cationic | Futaki and colleagues 2001 |
| LL-37 (comparison) | Human cathelicidin hCAP-18, residues 134-170 | LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES | 37 | +6 | Cationic amphipathic; not a CPP | Dürr and colleagues 2006 (review) |
The founding cell-penetrating peptides. Sequences and lengths are as given in the cited primary papers; net charge is counted from basic (Arg, Lys) and acidic (Asp, Glu) residues at neutral pH. LL-37 is included for the charge comparison only and is not a cell-penetrating peptide.
The table makes the point about LL-37 directly. Its net charge of +6, spread over 37 residues, is lower in density than TAT's +8 over 11, and its activity is at the membrane rather than across it. But the same residues that give a CPP its charge give LL-37 its charge, and they carry the same consequences for a lyophilised vial.
How uptake is described in cell models
The early papers described uptake that was rapid, happened at 4 °C, and did not saturate, all of which pointed to direct passage through the membrane rather than endocytosis. In 2003 Richard, Melikov, Vivès and colleagues re-examined this and found that much of the apparent uptake in fixed cells was an artefact: fixation redistributed peptide that had been bound to the cell surface into the interior. In living, unfixed cells, uptake of TAT and R9 was largely by endocytosis, slower and temperature-dependent, with a smaller direct component.
The current picture, as summarised by Guidotti and colleagues, is that both routes operate and the balance depends on the peptide, its concentration, the cargo attached and the cell type. At low concentration endocytic routes dominate; at higher concentration direct translocation becomes significant. Models proposed for the direct route include transient pore formation, inverted micelles and a carpet-like disruption of the bilayer, and these are models described in cell and artificial membrane systems rather than settled mechanisms.
Two cautions follow for anyone reading this literature as a buyer. Reported uptake efficiencies are highly method-dependent, and a figure from a fixed-cell fluorescence study is not comparable with one from live-cell flow cytometry. And cargo matters: a CPP that carries a small fluorophore efficiently may carry a protein poorly.
What the class is used for
CPPs are laboratory tools for getting something across a membrane in a model system. The cargo can be a fluorescent dye, a nucleic acid, a peptide or a protein, attached covalently or by a cleavable linker. Guidotti and colleagues review the range of cargoes and model systems. In this catalogue's terms, a CPP is bought as a reagent for cell work in the same way a fluorescent probe is, and it is typically ordered to a specified sequence from a custom synthesis house rather than from a stock list.
Why cationic peptides share handling problems
A peptide with eight positive charges in eleven residues behaves differently in a vial from a neutral one, and the differences are shared by any highly basic sequence, including LL-37. LL-37 explained covers that molecule; the handling points below apply across the group.
- Adsorption. Cationic peptides bind to negatively charged surfaces, which includes untreated glass and many plastics. A solution of a cationic peptide can lose a meaningful fraction of its content to the walls of a tube, and the loss is worst at low concentration. Low-binding labware and the inclusion of a carrier are standard mitigations in the cell-biology literature.
- Salt sensitivity. The electrostatic interactions that drive membrane binding are screened by salt. Activity measured in low-ionic-strength buffer will not be reproduced in physiological saline, and the same screening affects aggregation and solubility.
- Counterion load. A peptide purified by reversed-phase HPLC with trifluoroacetic acid is isolated as a TFA salt, with roughly one trifluoroacetate for each positive charge. For TAT that is eight counterions on an 11-residue chain, and the trifluoroacetate can be a quarter or more of the gross mass. TFA and acetate salt forms covers why that matters and when acetate exchange is worth specifying.
- Solubility. Highly charged peptides are generally very water-soluble, which is an advantage, but amphipathic ones such as penetratin can aggregate in water and may need a small amount of organic co-solvent. Peptide solubility explained gives the approach by sequence.
Ordering a CPP: what to specify and check
Because CPPs are custom rather than stock, the buyer writes the specification. The sequence should be stated in full, N to C, with the termini specified: TAT is usually ordered with a free N-terminus and a C-terminal amide, and penetratin likewise, but the choice changes the net charge by one and must be deliberate. Any label or cargo attachment point should be named, typically an added cysteine or lysine. The counterion should be specified, acetate if the downstream work is sensitive to trifluoroacetate. Purity should be stated as a minimum HPLC area percentage and identity confirmed by mass spectrometry, with the certificate reconciling the observed mass against the calculated mass for the specified termini and counterion-free peptide.
That is the same certificate discipline the catalogue's own guide to reading a certificate sets out for stock products, applied to a sequence the buyer chose. The chemistry is no different; only the responsibility for stating it has moved.